approaches five times the diameter of the hole, the strain becomes insensitive to the size of the erosion. This may indicate that erosion far from the defect does not influence the repair performance. 12.4.3 Hydrostatic Pressure Testing Ten hydrostatic tests were performed on damaged and repaired substrates. Nine for eroded holes and one for drilled hole. For these tests only four valid interface failures were observed, three for the eroded flaws and one for the drilled hole. This low percentage of valid tests may be due to the fact that the repair is very thin, so the failure mechanism changes from interfacial to through-repair pin holes or this could be the result of the reduction in interface strains predicted by FEA. Reducing the interface strains could shift the failure mode from an interfacial failure to a bulk composite failure. The average of the valid tests for eroded defects was 14 2.15 MPa. Plotting this value along with experimental data for drilled holes taken from the literature, it was observed that eroded flaws were below the trend [6]. However, a single experimental data point obtained for a drilled hole with the same through-wall damage size as the eroded flaw matches well the average value obtained for eroded holes (Fig. 12.9). This seems to indicate that fracture energy does not have any significant variation from drilled to eroded flaws. Adding in all tests regardless of failure mode, the average failure pressure is significantly less than would be expected from the drilled specimens. Further studies are underway to determine if the through-repair failures are the result of the thin wrap or a more systematic change in failure mode. 12.4.4 Digital Image Correlations Figure 12.10 shows the out of plane displacement obtained from the DIC analysis. As expected, the largest displacement occurs near the center of the defect. The DIC data was also compared to the theoretical displacement profile used by ASME and ISO given in Eq. (12.2) above. From this comparison, it is observed that the predicted profile significantly overpredicts the measured displacement profile. This overprediction will tend to lead to a more conservative design as any increase in pressure will lead to a larger increase in volume. The larger volume increase will, in turn, lead to higher driving energies for the crack, causing designers to apply thicker repairs as a result. Further study of the theoretical and experimental determined deformation profiles is ongoing. Fig. 12.9 Experimental data for failure pressure for drilled holes and eroded flaws 112 O. Ramirez and M.W. Keller
RkJQdWJsaXNoZXIy MTMzNzEzMQ==